中文摘要：

针对流域非点源污染的关键源区，进行最佳管理措施（BMPs）的配置，是非点源污染控制的有效途径。污染削减效率的准确识别对于BMPs在目标流域内的有效实施具有非常重要的意义。通过综合对比和分析实地监测、养分平衡、风险评估以及模型模拟等四类最佳管理措施评估方法的有效性、特点、适用条件及其局限性，得出以下结论：养分平衡法较为简便且易于使用，相较于其他方法，所需时间短且又可以消除评估效果的滞后效应，但对污染物削减的时间效应和传输过程影响考虑较少。风险评估和模型模拟方法可以更好地应对不同时空尺度下削减措施效率的评估，但需要大量实测数据的支持，同时模型模拟中普遍存在的时空不确定性影响很难消除。由于各种评估方法都有一定的适用条件，单一方法难以有效地完成评估目标，需要综合应用各类方法，才能最大程度地发挥这些方法的潜在功能和有效性，进而实现BMPs措施使用的成本一效益目标。

英文摘要：

The nonpoint source pollution can be controlled by implementing various best management practices （BMPs） in the watershed. However, before such practices are adopted, their effectiveness at various spatial and temporal scales must be evaluated. Though the effectiveness of individual BMPs has been usually assessed in standard plots, it is necessary to quantify the impact at a wider range （eg. at the watershed scale） to ensure that practices taken will be sufficient to meet EPL （the environmental protection law） targets. In this paper, we compare the characteristics and suitability of different approaches （direct measurement, nutrient budgeting, risk assessment and model simulation） to assess the effectiveness of actions to mitigate sources and transport of nitrogen （N） and phosphorus （P） from agricultural land to water. Nutrient budgets are most commonly used to quantify nutrient management by evaluating inputs and outputs over a defined time period. Automatic calculations under spreadsheets or other user friendly interfaces are used as accounting procedures to obtain the nutrient surplus or deficit values. System boundaries are flexible, resulting in a range of methodologies applicable from plot to national scale including farm gate, soil surface and soil system budgets. The limitation of this method that it currently fail to consider the timing and transport aspects of mitigation and assume a direct causal relationship between potential and actual nutrient loss likelihood of nutrient availability Risk assessment procedures quantify the risk of nutrient loss occurring based on the and delivery processes coinciding. Small areas may contribute disproportionately large amounts of nutrients and, by ranking vulnerability to loss, mitigation can be targeted to those areas at highest risk. The approach is most applicable to P loss which is explained by annual variations in runoff and the associated erosion of P enriched soils. But assessments demand increased availability of data, and there is a